Interactions among oligochaetes and a myxozoan parasite, Myxobolus cerebralis by Leah Candace Steinbach

Transcription

1 Interactions among oligochaetes and a myxozoan parasite, Myxobolus cerebralis by Leah Candace Steinbach A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences Montana State University Copyright by Leah Candace Steinbach (2003) Abstract: Myxobolus cerebralis, the causative agent of salmonid whirling disease, has caused the decline of two populations of wild rainbow trout {Oncorhynchus mykiss) on the Madison River MT and the Colorado River CO. The life cycle of Myxobolus cerebralis alternates between two hosts: various salmonid species and one aquatic oligochaete species, Tubifex tubifex. The objectives of this study were to determine the threshold myxospore dose needed to achieve complete infection within Tubifex tubifex and the influence that interactions among oligochaetes and M. cerebralis had on oligochaete fitness and the success of M. cerebralis. I examined the threshold myxospore dose of two T. tubifex strains experimentally and using nine myxospore doses. I determined that a dose of 50 myxospores per individual T tubifex achieved complete infection in the susceptible strain of T tubifex. Interactions were examined in two response surface competition design experiments using two strains of T tubifex, two species of oligochaetes and various doses of myxospores. The interactions among strains of resistant and susceptible T tubifex were density dependent. The infection prevalence and the proliferation of the parasite (triactinomyxon production) within treatments containing resistant and susceptible individuals were comparable to treatments containing only susceptible individuals. The interactions among susceptible T tubifex and Limnodrilius hoffmeisteri and the infection prevalence of T tubifex were density dependent. Tubifex tubifex had greater infection prevalence in treatments with L hoffmeisteri, but had variable triactinomyxon production. In both interaction experiments, susceptible T tubifex had greater adult growth when interacting with resistant T tubifex or L. hoffmeisteri and when exposed to myxospores. This suggests that susceptible T tubifex may have benefited from a low dose of myxospores. The results of these experimental interactions must be now be contrasted with interactions in natural systems. Further, the next task is to characterize oligochaete communities by strain of T tubifex and species of oligochaete, which has the potential to increase our understanding of host-parasite dynamics.

2 INTERACTIONS AMONG OLIGOCHAETES AND A MYXOZOAN PARASITE, MYXOBOLUS CEREBRALIS by Leah Candace Steinbach A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences MONTANA STATE UNIVERSITY Bozeman, Montana November 2003

3 N37??-t3H5 APPROVAL of a thesis submitted by Leah Candace Steinbach This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. Dr. Billie Kerans (Signature) / / PO 3 Date Approved for the Department of Ecology Dr. Scott Creel (Signature) Date Approved for the College of Graduate Studies

4 STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a master s degree at Montana State University, I agree that the Library shall make it available to borrowers under the rules of the Library.. If I have indicated my intention to copyright this thesis by including a copyright notice page, copying is allowable only for scholarly purposes, consistent with fair use as prescribed in the U. S. Copyright Law. Only the copyright holder may grant requests for permission for extended quotation from or reproduction of this thesis in whole or in parts. Signature Date 2J* (?$?>

5 iv ACKNOWLEDGEMENTS This research was funded by the National Partnership on Management of Wild and Native Coldwater Fisheries and I am appreciative of their financial support for such ecological research. I would like to give thanks to my committee, Dr. Billie Kerans, Dr. Charlotte Rasmussen and Dr. Cathy Zabinski for their support and guidance. Thanks to Russell Elwell for keeping me smiling. Also everyone who contributed to the editing process: Dr. Bob Garrott, Eric Bergman, Katie Brown, Chelsea Cada, Patty Hernandez, Adam Messer, David Richards, Jessie Salix, Lew Stringer, Jeff Warren and Julie Zickovich. Thanks to Jasper Buchbauer for laboratory assistance. A much appreciated thanks to Julie Zickovich for her excellent laboratory assistance. Also, a most appreciative thanks to Cal Frasier for his assistance and help in the laboratory. And a special thank you to Dr. Billie Kerans for providing me with so much of her personal energy, effort and friendship.

6 TABLE OF CONTENTS V Page LIST OF TABLES...vii LIST OF FIGURES...viii ABSTRACT... ix 1. INTRODUCTION...I THE LIFE CYCLE OF MYXOBOLUS CEREBRALIS...3 TUBIFEX TUBIFEXx m MYXOBOLUS CEREBRALIS... 4 TUBIFEX TUBIFEX, LIMNODRIL US HOFFMEISTERI AND MYXOBOL US CEREBRALIS OBJECTIVE I : INTERACTIONS AMONG TWO STRAINS OF TUBIFEX TUBIFEX AND MYXOBOL US CEREBRALIS INTRODUCTION...12 METHODS...17 Experimental Procedures Molecular Analyses Statistical Analyses RESULTS DISCUSSION OBJECTIVES 2 AND 3: INTERACTIONS AMONG TWO TUBIFICID SPECIES AND A MYXOZOAN PARASITE, AND THRESHOLD INFECTION DOSE INTRODUCTION METHODS Experiment I: Threshold infection dose...44 Experiment 2: Interactions between oligochaetes RESULTS Experiment I : Threshold infection dose Experiment 2: Interactions between oligochaetes DISCUSSION CONCLUSIONS 64

7 TABLE OF CONTENTS - CONTINUED PAGE REFERENCES CITED APPENDICES Appendix A: ANOVA Tables from Chapter Appendix B: Total triactinomyxon production from Chapter Appendix C: Strain A and strain B progeny identified from Chapter Appendix D: Infection prevalence from Experiment I, Chapter Appendix E: Triactinomyxon production from Chapter

8 vfi LIST OF TABLES Table Page 2.1. Analysis of variance results on the effects of oligochaete combination and myxospore dose on biomass response variables : Triactinomyxon variables of total triactinomyxons produced per day, infection prevalence and triactinomyxons produced by an individual per day for each oligochaete combination Analysis of variance results of the effects of oligochaete combination and myxospore dose on biomass response variables... 51

9 viii LIST OF FIGURES Figure I. I Page The life cycle of Myxobolus cerebralis, The effects of Tubifex tubifex strains and density and Myxobolus cerebralis on the total biomass, adult growth and progeny biomass production Randomly amplified polymorphic DNA gel identifying each strain of T. tubifex and possible hybrid progeny The relationship between level of myxospore exposure and the percentage of Tubifex tubifex releasing triactinomyxons The effects of oligochaete species and density and Myxobolus cerebralis infection on total biomass, adult growth and progeny biomass production The effect of oligochaete density and composition on the proportion of Tubifex tubifex releasing triactinomyxons The average triactinomyxon production of an individual Tubifex tubifex...

10 ABSTRACT Myxobolus cerebralis, the causative agent of salmonid whirling disease, has caused the decline of two populations of wild rainbow trout {Oncorhynchus mykiss) on the Madison River MT and the Colorado River CO. The life cycle of Myxobolus cerebralis alternates between two hosts: various salmonid species and one aquatic oligochaete species, Tubifex tubifex. The objectives of this study were to determine the threshold myxospore dose needed to achieve complete infection within Tubifex tubifex and the influence that interactions among oligochaetes and M. cerebralis had on oligochaete fitness and the success of M. cerebralis. I examined the threshold myxospore dose of two T. tubifex strains experimentally and using nine myxospore doses. I determined that a dose of 50 myxospores per individual T tubifex achieved complete infection in the susceptible strain of T tubifex. Interactions were examined in two response surface competition design experiments using two strains of T. tubifex, two species of oligochaetes and various doses of myxospores. The interactions among strains of resistant and susceptible T tubifex were density dependent. The infection prevalence and the proliferation of the parasite (triactinomyxon production) within treatments containing resistant and susceptible individuals were comparable to treatments containing only susceptible individuals. The interactions among susceptible T tubifex and Limnodrilius hoffmeisteri and the infection prevalence of T tubifex were density dependent. Tubifex tubifex had greater infection prevalence in treatments with L. hoffmeisteri, but had variable triactinomyxon production. In both interaction experiments, susceptible T tubifex had greater adult growth when interacting with resistant T tubifex or L. hoffmeisteri and when exposed to myxospores. This suggests that susceptible T. tubifex may have benefited from a low dose of myxospores. The results of these experimental interactions must be now be contrasted with interactions in natural systems. Further, the next task is to characterize oligochaete communities by strain of T. tubifex and species of oligochaete, which has the potential to increase our understanding of host-parasite dynamics.

11 I CHAPTER I INTRODUCTION Myxobolus cerebralis (Myxozoa: Myxosporea/ Actinosporea), the cause of salmonid whirling disease (El-Matbouli et al. 1992), was discovered in 1893 in Germany (Nickum 1996). Initially M cerebralis was a serious pathogen of hatchery trout, however modifications in hatchery practices reduced its damaging effects. Throughout the 1980 s, it was widely accepted that the parasite was not a threat to wild trout populations (Nickum 1996). However, recent evidence suggest that M cerebralis has caused declines of wild trout {Oncorhynchus mykiss) populations in the Intermountain West (USA) (Nehring and Walker 1996, Vincent 1996). The life cycle of M cerebralis involves two obligatory hosts: various salmonid species and an aquatic oligochaete, Tubifex tubifex (Oligochaeta. Tubificidae)(Wolf and Markiw 1984). Because the ecology of the salmonid and T. tubifex hosts are linked by M cerebralis, investigations of both hosts are required to fully understand interaction dynamics among all three species (Kerans and Zale 2002, Gilbert and Granath 2003). Studies examining the interactions between salmonids and M. cerebralis have revealed important facets of the whirling disease problem. For example, infection in salmonids varies with species, age and environmental variables (Halliday 1973, Markiw 1992, Hedrick et al. 1998, Zendt and Bergersen 2000, Ryce et al. 2001, Krueger 2002, MacConnell and Vincent 2002, Ryce 2003). Research on M cerebralis has focused on

12 the life cycle, pathology, phylogeny, and conditions that are favorable for the parasite (El-Matbouli at al. 1992, Siddall et al. 1995, El-Matbouli and Hoffmann 1989, El- Matbouli et al. 1999a, Kent et al. 2001). Research on the effects of M. cerebralis on T. tubifex has revealed infections are persistent, cause reduced growth and reproduction, and strains of T. tubifex vary in their susceptibility to infection (Gilbert and Granath 2001, Stevens et al. 2001, Kerans et al. submitted). Comparatively little research has been conducted on the interactions among oligochaetes and M. cerebralis, although such interactions may influence salmonid disease ecology in natural populations. The associations between two species are often affected by the presence of a third species, which may ultimately influence community structure (Price et al. 1986, Miller and Kerfoot 1987, Wootton 1994). Oligochaete interactions that may influence whirling disease ecology include interactions between and among strains of T. tubifex, and interactions between T. tubifex and other oligochaetes such as Limnodrilus hoffmeisteri (Stevens et al. 2001, Kerans et al. submitted). Little is known about the characteristics of aquatic communities that may be related to catastrophic infection in salmonids. It seems reasonable that the probability of infection of a salmonid will depend on a variety of abiotic and biotic factors, including interactions among hosts. Ifoligochaete assemblages and thus, interactions among oligochaetes differ appreciably among streams with high and low salmonid whirling disease severity, such interactions may influence salmonid infection. By characterizing the interactions among oligochaetes, I hope to make a connection between the mechanisms of infection among salmonid populations exposed to M. cerebralis. Our understanding of interactions

13 3 and their role in the success of M. cerebralis may improve the difficult task of predicting whirling disease incidents and understanding their influence on whirling disease ecology. The life cycle of Mvxobolus cerebralis Originally, M. cerebralis was thought to infect only salmonid hosts; however, further research proved that T. tubifex was also an obligatory host (Markiw and Wolf 1983, Wolf et al. 1986, El-Matbouli and Hoffmann 1989). Myxobolus cerebralis has two distinct spore stages: the triactinomyxon (TAM) and the myxospore. Genetic analyses confirmed that the alternate stages of the parasite belong to a single species (Andree et al. 1997) and revealed little genetic variation between M cerebralis from Europe and North America (Andree et al. 1999). The life cycle of M. cerebralis (Fig. 1.1) begins as T. tubifex ingests the myxospore stage while consuming sediments of aquatic substrate. Once the myxospore is ingested and inside the lumen of the intestine, it attaches itself to the intestinal mucosa. There, the germ cell moves between cells of the intestinal epithelium and reproduces asexually and sexually. After approximately days, depending on T. tubifex strain and temperature (El-Matbouli and Hoffmann 1989, Stevens et al. 2001, Kerans et al. submitted, B. L. Kerans and R. I. Stevens, Montana State University, unpublished data), the oligochaete releases a buoyant stage of the parasite into the water, the TAM. The TAM floats in the water and, upon contact with the epidermis of a salmonid; the sporoplasms of the TAM are injected. The parasite migrates through the epidermis to the nervous tissue and finally to the cartilaginous

14 4 skeletal system where it produces myxospores (El-Matbouli et al. 1995). The fish may also become infected if it consumes infected oligochaetes (Markiw 1986). As the infection progresses in the fish, cartilaginous regions are consumed, loss of function ensues and survivability decreases (Hoffman et al. 1962, Hedrick et al. 1998). Clinical signs of the disease include blackened areas of the flesh and skin, deformed skeletal structure and characteristic tail chasing whirling behavior (Hoffman et al. 1962). Once the fish dies, the decaying carcass releases M cerebralis myxospores into the sediment where T. tubifex may ingest myxospores. Tubifex tubifex and Mvxobolus cerebralis Tubifex tubifex is a widespread and environmentally tolerant oligochaete that is found in most aquatic habitats (Brinkhurst and Jamison 1971, Chapman et al. 1982). In monocultures, the growth and reproduction of T tubifex is density dependent suggesting intraspecific interactions influence population size (Kosiorek 1974, Bonacina et al. 1989) Growth rates, reproduction and environmental versatility vary among T. tubifex from different global regions, suggesting that the species is comprised of multiple genetic strains (Poddubnaya 1980, Milbrink 1983). Molecular analysis of populations of T tubifex has demonstrated genetic differences among individuals from various regions in North America and Europe (Anlauf 1994, 1997, Sturmbauer et al. 1999, Erseus et al. 2000, Beauchamp et al. 2001, Beauchamp et al. 2002, Kerans et al. submitted). Studies

15 5 of local populations of T. tubifex on the upper Colorado River, Colorado suggest T. tubifex are comprised of multiple strains (Beauchamp et al. 2002). myxospore triactinomyxon (TAM) Figure The life cycle of the parasite Myxobolus cerebralis. The parasite has two distinct transmission stages, the myxospore and the triactinomyxon, which are infective to the obligatory hosts Tubifex tubifex and many salmonid species, respectively.

16 6 The interactions between individual strains of T. tubifex and M. cerebralis have been examined. The effects of M. cerebralis on T. tubifex are negative and the strength of the effect is positively related to TAM production (Kerans et al. submitted). Susceptible strains of T. tubifex that produce high numbers of TAMs have lower fitness than uninfected susceptible individuals. Moderately susceptible T. tubifex that produce few TAMs have lower fitness than uninfected individuals (Stevens et al. 2001, Kerans et al. submitted). Some strains of T tubifex do not produce TAMs and have not been infected (Beauchamp et al. 2002). However, the fitness of resistant T. tubifex challenged with myxospores is unknown. The decrease in fitness among susceptible T. tubifex may be the result of increased mortality or decreased reproductive output due to parasitism. Tubifex tubifex exhibit low mortality if exposed to low doses of myxospores, and morality increases only with very high myxospore doses (Gilbert and Granath 2001, Stevens et al. 2001). Furthermore, Kerans et al. (submitted) suggest that reduced reproductive output of infected T. tubifex is the cause of fitness reduction. The effects of T. tubifex on M. cerebralis are primarily positive and appear directly related to the susceptibility of T. tubifex. Myxobolus cerebralis is successfully transmitted by susceptible and moderately susceptible T. tubifex that produce TAMs, whereas, the transmission success of M. cerebralis decreases with resistant T. tubifex that do not produce TAMs. Several factors may be responsible for the fact that resistant T. tubifex do not produce TAMs: myxospores may not be able to invade gut tissue, myxospores may invade but are destroyed by immune response or oligochaetes may have the capability of deactivating myxospores (M. El-Matbouli, University of Munich,

17 I personal communication). Therefore, the interactions among resistant and susceptible T. tubifex could reduce infection among susceptible individuals because resistant individuals may deactivate myxospores or reduce the overall TAM production because they consume myxospores. Plausibly, communities of tubificids containing strains of T. tubifex that vary in susceptibility could influence the success of M. cerebralis and therefore infection in salmonids. It is unknown how mixed communities of resistant and susceptible T. tubifex can affect the success of M. cerebralis or how M cerebralis affects the success of particular strains of T. tubifex. Consequently, it is important to genetically catalogue and understand oligochaete communities throughout areas where M. cerebralis has been detected (Beauchamp et al. 2002, Kerans et al. submitted). Interactions among strains of T. tubifex exposed to M. cerebralis are likely to differ from those not exposed to myxospores. I expect infected T. tubifex to have reduced growth and reproduction compared to uninfected T. tubifex, consistent with findings of Stevens et al I would expect the competitive ability of susceptible T. tubifex, when growing in combination with resistant T. tubifex, to be lower when susceptible oligochaetes were infected than when not infected. Further, if resistant T. tubifex can deactivate myxospores then its presence could reduce the infection prevalence among susceptible T. tubifex. Therefore, comparing and contrasting interactions of various strains of T. tubifex and their subsequent production of M. cerebralis may provide us with information of how infection levels can vary among T. tubifex within streams and potentially how this relates to salmonid infection. Experimental inter-strain interactions must then be compared to interactions in natural systems. Understanding oligochaete

18 8 interactions has the potential to improve our abilities to predict potential whirling disease outbreaks and disease severity in salmonids, and further characterize host-parasite interactions among varying host strains. Tubifex tub if ex ^ Limnodrilus hoffmeisteri and Mvxobolus cerebralis Many studies of tubificid ecology, conducted throughout the past century (e.g., Kennedy 1966, Chapman et al. 1982, Bonacina et al. 1989), have examined life history, tubificid response to environmental pollutants, population dynamics and interactions among tubificid species. Tubifex tubifex and Limnodrilus hoffmeisteri (Oligochaeta: Tubificidae), two oligochaetes that typically inhabit stream communities, have similar distributions, physical attributes, and functional roles but different ecological requirements (Brinkhurst and Jamieson 1971, Block et al. 1982, Mermillod-Blondin et al. 2001). For example, both oligochaetes are deposit feeders but each prefers different sediment sizes and depths within the substrate (Mermillod-Blondin et al. 2001). In monocultures, individual growth and reproduction of T tubifex and Limnodrilus sp. are density dependent suggesting intraspecific competition influences oligochaete populations (Kosiorek 1974, Poddubnaya 1980, Bonacina et al. 1989). Interactions between L hoffmeisteri and T tubifex are apparent in mixed cultures, where their respiration rates are lower and population growth rates increase compared to monoculture each species (Brinkhurst 1974, Milbrinlc 1987), suggesting a facultative

19 9 mutualism with each species providing the other with beneficial food resources (Milbrink 1987). However in other experiments, L hoffmeisteri depressed the growth of T tubifex when M. cerebralis was not present suggesting that nutrients or space may have been limiting factors for these mixed oligochaete cultures (B.L Kerans and J.C. Lemmon, Montana State University, unpublished data). Conversely, when M. cerebralis was present, T. tubifex grew much better in the presence of L. hoffmeisteri than in pure culture, suggesting facultative mutualism (B.L Kerans and J.C. Lemmon, Montana State University, unpublished data). The complex and variable interactions between T. tubifex and L hoffmeisteri warrant further investigation, particularly in their influence on M. cerebralis success. The occurrence of both positive and negative interactions between T. tubifex and L. hoffmeisteri suggest that the interactions can vary with biotic factors. Shifts between positive and negative interactions along environmental gradients have been documented among plants and marine invertebrates (e.g. Callaway and Walker 1997, Dudgeon and Petraitis 2001). Further, species interactions comprised of both competitive and facultative components can vary spatially and temporally (Bruno et al. 2003), depending on environmental conditions, indirect interactions, and the density and diversity of species (Callaway and Walker 1997). This suggests the importance of observing interactions under various conditions simultaneously to properly assess the types of interactions that may be occurring. Ifthe primary interaction between L. hoffmeisteri and M. cerebralis is mutualistic, I predict that the interactions among uninfected T tubifex and L hoffmeisteri to result in

20 10 enhanced growth of both species compared to growth in monocultures. As competitors, uninfected T tubifex and L. hoffmeisteri would have depressed growth compared to monocultures. As mutualists, infected T. tubifex compared to uninfected T. tubifex would have reduced growth and provide less benefit to L hoffmeisteri. Additionally, the growth of infected T. tubifex would be less than L. hoffmeisteri (Stevens et al 2001), due to decreased competitive ability of T. tubifex. Competitive interactions will result in lower growth of T. tubifex, than mutualist interactions. Because L. hoffmeisteri cannot be infected by M. cerebralis (Wolf et al. 1986, Kerans et al. submitted), it is possible that L. hoffmeisteri deactivates myxospores by digestion as it feeds. The feeding activity of L. hoffmeisteri may reduce myxospore abundance and alter the prevalence of infection in T. tubifex populations. Therefore, I would expect the growth of T. tubifex grown with L. hoffmeisteri to be less depressed compared to monoculture because L hoffmeisteri has deactivated myxospores and hence minimized the negative effects of M. cerebralis infection upon T. tubifex. Thus, the type of oligochaete interaction and the ability of L. hoffmeisteri to deactivate myxospores will influence the success of M. cerebralis. For example, the success of M. cerebralis will be greatest in a community where T. tubifex is present and L hoffmeisteri does not deactivate myxospores. The objectives of my work were to: I. Examine the interactions among strains of Tubifex tubifex and Myxobolus cerebralis.

21 11 2. Examine the interactions among Tubifex tubifex, Limnodrilus hojfmeisteri and Myxobolus cerebralis. 3. Determine the threshold dose of myxospores to achieve complete infection among Tubifex tubifex. The purpose of objectives one and two were to determine the relative importance of interactions on the success of T tubifex and M. cerebralis transmission. Each is addressed in a separate chapter. Objective three determined the doses of myxospores I used in the other objectives and is addressed in Chapter 3. The final chapter, Conclusions, integrates the various sections, provides suggestions for applying this work in management situations and possible future research. This research will hopefully reveal how oligochaete interactions influence M. cerebralis success and potentially provide information for predicting future M. cerebralis infections based on oligochaete communities.

22 12 CHAPTER 2 OBJECTIVE I: INTERACTIONS AMONG TWO STRAINS OF TUBIFEX TUBIFEX AND MYXOBOLUS CEREBRALIS Introduction The dynamics of host-parasite interactions may influence host population growth and host evolution (Anderson and May 1979, May and Anderson 1979, May and Anderson 1990). The virulence of a parasite depends on the relative susceptibility of a specific host, parasite phenotype, host density and competition among parasites within hosts (Ewald 1983, Bull 1994). Moreover, the evolution of host susceptibility and parasite success depends on the variability of virulence among host-parasite combinations (Vandame et al. 2000). Further, the variation among host susceptibility and parasite virulence may affect the outcome of interactions among hosts (Ebert et al. 1998, Yan et a l.1998). Parasite-mediated competition among hosts was first noted in Park s (1948) classical experiment with two species of tribolium beetle and a protozoan parasite, showing that a parasite could reverse the outcome of competition between two beetle species. Later research supports the importance of parasite-mediated interactions among competitors (e.g., Settle and Wilson 1990, Grosholz 1992, Schall 1992, Yan et al. 1998).

23 13 Models of parasite-mediated competition predict that in the presence of a parasite, hosts with low susceptibility will be superior competitors to hosts with high susceptibility (Yan 1996). However, interactions among hosts depend on the environment and host life history parameters, such as geographic range, population density, growth rate and body size (Price et al. 1988, Begon et al. 1992, Bruno et al. 2003). For example, when a parasite common to two host species has a differential impact upon host fitness, the host species with a wide geographic range will displace host species with a narrow range because of population size and associated immune response (Price et al. 1988). The geographic-range hypothesis assumes that host species with a large geographic range will have a higher effective population size than the host species with a small geographic range. Consequently, the host with the larger geographic range and population size can maintain parasitic populations and stimulate rapid immune response (Price et al. 1988). I examined intra-strain and inter-strain variation in the response of Tubifex tubifex to parasitism by Myxobolus cerebralis. Myxobolus cerebralis (Myxozoa: Myxosporea /Actinosporea) causes whirling disease in salmonids (El-Matbouli et al. 1992) and alternates between two obligate hosts: various salmonid species and one known aquatic oligochaete species, Tubifex tubifex (Oligochaeta: Tubificidae) (Markiw and Wolf 1983, Wolf and Markiw 1984). At least two populations of wild rainbow trout (Oncorhynchus mykiss) within the Intermountain West (USA) have declined as a result of M. cerebralis (Nehring and Walker 1996, Vincent 1996). Myxobolus cerebralis from various locations in Europe and North America exhibit very little genetic variation (Andree et al. 1999).

24 14 Myxobolus cerebralis has a complex life cycle with two spore stages, the myxospore and the triactinomyxon (TAM). Tubifex tubifex ingests the myxospore stage while consuming sediments. Parasite reproduction occurs between epithelial cells of the intestine of T. tubifex. After approximately days, depending on T. tubifex strain and temperature (El-Matbouli and Hoffmann 1989, El-Matbouli et al. 1999a, Stevens et al. 2001, Kerans and Stevens in preparation), the buoyant stage of the parasite, the triactinomyxon (TAM), is released into the water column. The TAM floats in the water and when a salmonid is contacted, sporoplasms of the TAM are injected into the epidermis. The degradation of the cartilaginous skeletal system occurs as myxospores are produced (El-Matbouli et al. 1995). Myxospores are released into the sediment after infected salmonids die and their carcasses decompose. Tubifex tubifex is a widespread and environmentally tolerant oligochaete that is found in most aquatic habitats (Brinkhurst and Jamison 1971, Chapman et al. 1982). In monocultures, the growth and reproduction of T tubifex is density dependent suggesting intraspecific interactions influence populations (Kosiorek 1974, Bonacina et al. 1989). However, populations of T tubifex are distinguished by genetic variation (Sturmbauer et al. 1999, Beauchamp 2001). Strains of T tubifex differ in reproduction, growth rates and environmental versatility (Poddubnaya 1980, Milbrink 1983). Further, molecular analysis has demonstrated genetic differerices among individual T tubifex from North America and Europe (Anlauf 1994,1997, Sturmbauer et al. 1999, Erseus et al. 2000, Beauchamp et al. 2001, Beauchamp et al. 2002, Kerans et al. submitted) and studies of

25 15 local populations of T. tubifex in the Colorado River CO showed that T. tubifex populations are comprised of multiple strains (Beauchamp et al. 2002). The effects of M. cerebralis on T. tubifex are negative and the strength of the effects is related to TAM production (Stevens et al. 2001, Kerans et al. submitted). Some strains of T. tubifex do not produce TAMs and cannot be infected (Beauchamp et al. 2002). Further, the fitness of resistant T. tubifex challenged with myxospores is unknown. The decrease in fitness among susceptible T. tubifex may be the result of increased mortality or decreased reproductive output. Tubifex tubifex exhibit low mortality if exposed to low doses of myxospores, and morality increases only with very high myxospore doses (Gilbert and Granath 2001, Stevens et al. 2001). Furthermore, Kerans et al. (submitted) suggest that reduced reproductive output is the main cause of fitness reduction. The effects of T. tubifex on M. cerebralis are primarily positive, depending on the susceptibility of T. tubifex. Myxobolus cerebralis is successfully transmitted by susceptible and moderately susceptible T. tubifex that produce TAMs, whereas the transmission success of M. cerebralis decreases with resistant T. tubifex that do not produce TAMs. Resistant T. tubifex that do not produce TAMs suggest immunity to myxospore invasion or a capability to deactivate myxospore prior to invasion (M. El- Matbouli, University of Munich, personal communication). Therefore, the interactions among resistant and susceptible T. tubifex could reduce infection among susceptible individuals because resistant individuals deactivate myxospores or reduce the overall TAM production because they consume myxospores. Plausibly, tubificid communities

26 16 that vary in susceptibility could influence the success of M. cerebralis and therefore infection in salmonids. It is unknown how mixed communities of resistant and susceptible T. tubifex affect the success of M. cerebralis or how M. cerebralis affects the success of particular strains of T. tubifex. Consequently, it is important to genetically catalogue and understand oligochaete communities throughout areas where M. cerebralis has been detected (Beauchamp et al. 2002, Kerans et al. submitted). I conducted a laboratory experiment that examined the interactions among susceptible and resistant T. tubifex and M cerebralis. The major objectives of this research were to determine: I) whether the interactions between strains of T. tubifex differed in strength or direction when they were exposed and not exposed to M. cerebralis, 2) whether resistant T. tubifex influenced the transmission of M. cerebralis by susceptible T. tubifex, and 3) whether resistant T. tubifex can be infected when exposed to high doses of myxospores. I predict susceptible and resistant T tubifex would exhibit intra-strain and interstrain density dependent competition. Further, the addition of M. cerebralis would not affect intra-strain interactions. However, I predict parasitized, susceptible T. tubifex to have reduced growth and reproduction compared to unparasitized, susceptible T. tubifex or resistant T. tubifex. With the addition of resistant T. tubifex, susceptible T. tubifex would have reduced infection compared to susceptible T. tubifex in monoculture due to deactivation of myxospores by resistant T. tubifex. Further, I predict resistant T tubifex would become infected when exposed to extremely high doses of myxospores. Finally, I predict the success of M cerebralis to be greater in treatments with only susceptible T.

27 17 tubifex compared to treatments with both resistant and susceptible T tubifex because of greater production of TAMs by susceptible individuals. Methods Experimental procedures I used two strains of T. tubifex: strain A that had produced large numbers of TAMs in previous experiments (Stevens et al. 2001, Kerans et ah submitted) and strain B that had been reported as not susceptible to M. cerebralis infection (Beauchamp et al. 2002), both from disease-free, pure laboratory cultures. Strain A was collected from the Mt. Whitney California State fish hatchery settling ponds, where M. cerebralis has been enzootic since 1984 (Modin 1998). Strain B was obtained from bays of Lakes Erie and Ontario where M. cerebralis was first documented in the 1950 s in wild salmonids (Reynoldson et al. 1996, R. Nelson, United States Fish and Wildlife Service, La Crosse Wisconsin, personal communication). These cultures have been maintained in our laboratory since 1997 (see Stevens et al. 2001, Kerans et al. submitted, for detailed culture propagation methods). Myxobolus cerebralis myxospores were extracted from infected rainbow trout using continuous centrifuge methods (O Grodnick, 1975). This process results in an emulsion containing myxospores. Trout originated at a Montana State fish hatchery, Ennis MT and were infected in our laboratory with TAMs from the United States Fish

28 18 and Wildlife Service, Fish Health Laboratory, Bozeman MT. Using I jal of extracted myxospore emulsion, I determined the abundance of myxospores by counting all viable myxospores on a hemacytometer with a compound microscope (400X magnification). The total number of myxospores in the emulsion was extrapolated from the mean number of myxospores from four hemacytometer counts and was used to create the required experimental doses. The emulsion was evenly mixed with a stirring plate as myxospores were removed with an automated pipette and administered to oligochaetes. The laboratory experiment examined the intra-strain and inter-strain interactions between strain A and strain B T tubifex when myxospores were present and absent. Fifty T. tubifex A, 100 T tubifex A, 50 T. tubifex A + 50 T. tubifex B, 50 T. tubifex B and 100 T. tubifex B were exposed to 0 (3 replicates) or 2500 (4 replicates) myxospores per container, in a response surface competition design (Goldberg and Scheiner 2001). Oligochaete densities were within the natural range found in streams (6,000 to 13,000 individuals/m2) (Lazim and Learner 1986). Experimental densities of 50 and 100 individuals per container translate to natural densities of 6180 and 12,360 individuals/m2. I chose the dose of 2500 myxospores per container because it is below the threshold of 100 % infection (See Chapter 3, Results), which allowed me to observe interactions between T. tubifex when myxospores were abundant enough to cause high infection prevalence, but not so abundant that the effects of the interactions would be obscured due to excess myxospores. A response surface design allowed me to partition and evaluate intra-strain and inter-strain interaction effects. For example by comparing measures of fitness of each strain between densities of 50 and 100,1 can examine intraspecific

29 19 interactions. Alternatively, by comparing measures of fitness between densities of 100 strain A and 50 strain A + 50 strain B, I can examine inter-strain interactions. Additionally, I prepared one treatment with three replicates of strain B at a density of 100 that were exposed to 50,000 myxospores per container to infect resistant strain B T. tubifex and to examine how high doses affect the fitness of this strain. At the beginning of the experiment, oligochaetes were removed from mass cultures and were placed in plastic containers for 24 hours without substrate to equalize hunger levels. I then wet weighed T. tubifex in groups of 50 to the nearest gram. The appropriate numbers of oligochaetes were placed in plastic containers (9x9x5 cm) with 40 ml of sand, I ml of organic matter, and 200 ml of dechlorinated tap water and inoculated with myxospores. Treatments without myxospores were inoculated with a similar quantity of emulsion from parasite-free rainbow trout. Oligochaetes were exposed to myxospores for 5 days in accordance with previous myxospore exposure procedures (see Chapter 3). Thereafter, all oligochaetes were removed from each container and placed in a new container with fresh substrate and water. I maintained containers, supplied with a constant flow of air, in incubators at 15 C with a 12:12 h L:D photoperiod. Oligochaetes were fed 0.10 to 0.12 grams of dried Spirulina weekly, which is sufficient for growth and reproduction (J. C. Lemmon, Montana State University, personal communication). I replaced approximately 95 % of the water from each container with fresh dechlorinated water weekly. Prior to expected TAM production, on day 65 post-exposure to myxospores, I began to scan for TAMs in each replicate. The water from each container was poured

30 20 separately through a 20 pi sieve that retained TAMs. Triactinomyxons were back flushed and concentrated in 10 ml of water. I scanned water from two, 55 pi aliquots with a phase contrast microscope (IOOX magnification) every three days until TAMs were found. Once TAMs were detected, I began counting TAMs weekly. I did TAM counts on all replicates until day 108 post-exposure. After day 108 post-exposure, I counted TAMs from two replicates for each myxospore positive treatment and all three replicates for each treatment that were not exposed to myxospores (see below). Average TAM production per replicate was recorded as the mean number of TAMs counted from two slides each week, which was extrapolated to determine the quantity of TAMs contained in the TAM concentrate. Total TAM production was calculated by summing the average TAM production per replicate per week. On day 108 post-exposure, two replicates from each treatment that were exposed to myxospores were randomly chosen to determine the prevalence of infected T. tubifex. Adult oligochaetes were removed from plastic containers and placed individually in 12- well plates, with 4 ml capacity. After two days, to allow TAMs to accumulate in the wells, I scanned plates with a dissecting microscope (I OX magnification) and recorded individuals as positive if TAMs were visible in the well. I scanned plates two times within a 10-day period, each time replacing the wells with fresh dechlorinated water. The prevalence of infected T. tubifex was calculated for each replicate by dividing the number of individuals producing TAMs by the total number of T. tubifex observed. I measured three response variables for each strain of T. tubifex to assess TAM production and

31 21 infection prevalence of M. cerebralis. These response variables were based on equation (2 1). Total TAM production / days observed = [susceptible strain A (infection prevalence)(initial density)(total TAMs produced / {number of infected individuals*days observed}) + resistant strain B (infection (2 I) v ' ' prevalence) (initial density)(total TAMs produced / {number of infected individuals*days observed})] Total TAM production per day was calculated by dividing the total TAMs produced per replicate by the days observed (hereafter referred to as TAM production ). Total TAMs produced per infected individual per day was calculated by dividing the total TAMs produced by the number of infected individuals and the days observed (hereafter referred to as TAMs per individual ). Prevalence of infection was calculated by dividing the total number of individual adults producing TAMs by the number of adults observed. Triactinomyxon response variables from the 50 strain A + 50 strain B treatments were calculated based on a density of 50 because the molecular assay showed that only T. tubifex strain A produced TAMs (See Results). At the termination of a replicate (either day 108 or day 135 post-exposure), all adults and progeny were weighed to the nearest gram. Further, because strains of T. tubifex are physically identical, all adults from 50 A T. tubifex + 50 B T tubifex treatments were analyzed with molecular markers to identify the strain of each individual (see Molecular Analyses). Further, fifty or thirty randomly chosen progeny from each

32 22 replicate were analyzed using molecular markers to determine which strain of progeny were produced. The effects of host-parasite interactions are often associated with reduced growth and reproductive success (Forbes 1993). I measured three responses to assess the effects of interactions among strains of T tubifex and M. cerebralis on the fitness of the oligochaetes: I) total biomass production per initial individual per day (((final adult weight -initial adult weight) + final progeny weight) / initial density of adults*days of growth) (hereafter referred to as total biomass ), 2) adult growth per initial individual per day (((final adult weight / final density) - (initial adult weight / initial density of adults))/days of growth) (hereafter referred to as adult growth ), and 3) progeny biomass produced by an individual adult per day (final progeny weight / initial density of adults*days of growth)(hereafter referred to as progeny biomass ). I separated total biomass into response variables of adult growth and progeny biomass to determine how each factor influenced total biomass. Each set of response variables was measured for each strain of T. tubifex separately. Because significant mortality typically occurs only when myxospore doses are high (Stevens et al. 2001), I did not examine oligochaete mortality as a response variable, and mortality during this experiment was 2 % (± I % SD). I calculated mortality by subtracting the abundance of oligochaetes at the start from the abundance of oligochaetes at the end in each replicate and averaged among all replicates. Oligochaete strain-density combinations will hereafter be referred to as oligochaete combination.

33 23 Molecular Analyses To identify the strain of each T. tubifex within 50 strain A + 50 strain B treatments, I used a maximum of three randomly amplified polymorphic DNA (RAPD) primers that provided ample banding differences between the strains. Genomic DNA was extracted from whole adult and progeny T. tubifex using Nucleospin Multi-96 tissue kits as per BD Biosciences Clontech protocol (PT3629-1) and was frozen until needed. I assessed the quantity and quality of DNA using gel electrophoresis with 5.0 pi of sample DNA on 1.5 % agarose gels with 0.5 X Tris-acetate (TAB) buffer. Two types of DNA ladders were used to compare DNA samples: 0.5 pi each of IKb and hyperladder I Bioline. Ethidium bromide was used to stain gels and visualize the DNA bands with ultra-violet light. All gels were documented using Lab Works program with a UVP EPI-Chemi dark room. Only one RAPD primer was necessary to compare and assess the genetic composition from polymerase chain reaction (PCR) amplified adult T tubifex DNA. The RAPD primer consists of the following base pairs; 5'-GGG CCG TTT A-3 (see Rasmussen et al for details of RAPD primer development). The PCR reaction mixture contained the following standard reagents: Sigma IX buffer without MgCl2, 2.5 mm MgCl2,200 um dntp s, 20 picomoles of RAPD primer, Research Genetics Inc. Rediload, 1.0 unit of Sigma Taq DNA polymerase and either 0.5 or 1.0 pi of sample DNA. All PCR amplification followed the subsequent procedure on a MJ Research, Inc. PT-100 thermocycler (Watertown, MA): I) 930C for 3 minutes, 2) 930C for 15 seconds.

34 24 3) decrease by 47 C degrees at a rate of 0.4 C per second, 4) 46 C for I minute 30 seconds, 5) increase by 26 C at a rate of OA0C per second, 6) 72 C for I minute 30 seconds, 7) repeat steps 2 through 6, 34 times, 8) hold at 15 C. Following PCR amplification, products were visualized by electrophoresis on 2% 0.5 X TAE agarose gels. Each strain of T. tubifex from 50 strain A T. tub if ex + 50 strain B T. tubifex treatments was distinguished by different banding patterns from PCR products. All individuals were allocated to a strain and the proportion of each strain was calculated by dividing the number of each strain identified by PCR by the total number of T. tubifex identified. The proportion of each strain could then be applied to biomass calculations. To determine the strain of each progeny, I analyzed individual progeny using three RAPD primers with the following sequences; #18: 5 -GGG CCG TTT A-3, #2: 5 - CCT GGG CTT G-3, and #54: 5 -GTC CCA GAG C-3. Three primers were used to reduce the possibility of not recognizing a hybrid progeny. Considering the number of banding pattern differences (at least two between each strain among each primer), I had a I in 8 (1A x 1Z2 x 1Z2) chance of not recognizing a hybrid progeny. I calculated the relative proportion of each strain by dividing the identified strain number by the total number of progeny amplified which was then applied to calculate the total number of progeny produced per strain.

35 25 Statistical analyses Because replicates used for infection prevalence had a shorter growth period, biomass response variables were tested for differences between the days of growth, (108 or 135 days) and among oligochaete combinations using two-factor analyses of variance (PROC GLM, SAS Institute, 2001). Total biomass, adult growth or progeny biomass production did not differ between days of growth, oligochaete combinations or with the interaction between the two factors (Appendix A). Therefore, data were combined and each biomass variable was tested for significant differences among oligochaete combination and myxospore dose using two-factor analyses of variance (ANOVA, SAS Institute, 2001). The treatment of 100 strain B T. tubifex with 50,000 myxospores was not used in the calculation of the biomass response variables of strain B because their inclusion created imbalance in the ANOVA analyses model. Residuals were examined for normality and variances stabilized with log or square root transformations, where necessary. To examine the fitness of strain B at high doses, I compared total biomass, adult growth and progeny biomass of 100 strain B exposed and not exposed to myxospores using a t-test for significant differences between myxospore doses of 0 and 50,000 (SAS Institute, 2001). Because two replicates from each treatment were removed from the daily calculation of TAM production on day 108, total TAM production was initially tested for differences in the number of days observed producing TAMs (end day of 108 or 135) and among oligochaete combinations using two-factor ANOVA (SAS Institute, 2001).

36 26 Triactinomyxon production did not differ among oligochaete combinations, days observed nor with the interaction between the two factors (Appendix A). Therefore, all replicates were combined for analyses. Triactinomyxon production, TAMs produced per individual and prevalence of infected T tubifex were tested for significant differences among oligochaete combinations using one-way ANOVAs (SAS Institute, 2001). The response variables were examined for normality and did not require transformation. Tulcey s Honestly Significant Differences tests (Tukey s HSD) were used in all analyses to determine differences among treatment levels when main effects were significant (P < 0.05). Results The total biomass of strain A was highest at densities of 50 individuals and decreased as densities increased to 100 with the addition of either strain of T. tubifex. (Table 2.1, Tukey s HSD, P < 0.05, Fig 2.1a). Total biomass did not differ between myxospore doses, nor with the interaction between oligochaete combination and myxospore dose (Table 2.1). The adult growth of strain A was highest at densities of 50 alone or when in combination with 50 strain B, and lowest in 100 strain A treatments (Table 2.1, Tukey s HSD, P < 0.05, Fig 2.1b). Growth of adults was greatest when exposed to myxospores (Table 2.1, Fig 2.lb). The interaction between oligochaete combination and myxospore dose was not significant.

37 27 Table Analysis of variance results for the effects of oligochaete combination and myxospore dose on biomass response variables. All biomass response variables are expressed as a per capita daily rate. Response Oligochaete variable strain* (f) Source of variation ss F df P Total biomass A dose I 0.62 oligochaete combination < interaction error B dose I 0.75 oligochaete combination < interaction error Adult growth A (sqrt) dose 2.16E I 0.01 oligochaete combination 2.65E interaction 8.13E error 4.36E B(In) dose 1.47E I 0.67 oligochaete combination 4.59E interaction 8.64E error 1.19E Progeny biomass A dose I 0.23 oligochaete combination < interaction error B dose I 0.22 oligochaete combination < interaction error * A represents strain A T tubifex, B represents strain B T. tubifex (t)=the transformation: (sqrt)= square root, (In)=Iog = sums of squares

38 28 Strain A T. tubifex produced greater biomass of progeny at densities of 50 individuals, and as densities increased to 100 with the addition of either strain, the progeny biomass decreased (Table 2.1, Tulcey s HSD, P < 0.05, Fig 2.1c). Progeny biomass did not differ between myxospore doses, nor with the interaction between the two factors (Table 2.1, Fig. 2.1c). The total biomass of strain B was highest at densities of 50 and lowest at densities of 100 individuals of strain B or in combination with strain A (Table 2.1, Tukey s HSD, P < 0.05, Fig 2.Id). Total biomass did not differ between myxospore doses, nor with the interaction between the two factors (Table 2.1, Fig 2.Id). However, when comparing the total biomass between 0 and 50,000 myxospore doses, the total biomass of strain B was greatest when exposed to 0 myxospores (Mean ± I SE: mg / individual*day ± 0.015, n = 3) and lower at a dose of 50,000 myxospores (Mean ± I SE: mg /individual*day ± 0.011, U ~ 4.95, P = , n - 3). The adult growth of strain B did not differ between oligochaete combinations, myxospore doses, nor with the interaction between the two factors (Table 2.1, Fig 2.1e). Furthermore, when comparing the adult growth between 50,000 and 0 myxospores, the adult growth of strain B did not differ (t< = 1.24, P = , n = 3). Strain B T. tubifex produced greater biomass of progeny at densities of 50 individuals and as densities increased to 100 with the addition of either strain the progeny biomass decreased (Table 2.1, Tukey s HSD, P < 0.05, Fig 2.If). Progeny biomass did not differ between myxospore doses, nor with the interaction between the two factors (Table 2.1). However, strain B produced greater progeny biomass when exposed to 0

39 29 myxospores (Mean ± I SE: mg / individual*day ± 0.014, n = 3) than when exposed to 50,000 myxospores (Mean ± I SE: mg / individual*day ± 0.006, t4 = 5.43, P = , n = 3). Strain A O alone, 0 myxospore alone, 2500 myxospore w/a, 0 myxospore w/a, 2500 myxospore 50 A 100 A 50A/50B Oligochaete Combination 50 B 100 B 50A/50B Oligochaete Combination Figure The effects of Tubifex tubifex strain, density and relative abundance of the two strains and Myxobolus cerebralis infection on biomass variables. Each column illustrates the particular variable for the target strain listed, a) total biomass, b) adult growth, c) progeny biomass, d) total biomass, e) adult growth, and f) progeny biomass. Error bars are one standard error of the mean.

40 30 I determined that individuals producing TAMs in the 50 strain A + 50 strain B treatments were all strain A T. tubifex using a randomly amplified polymorphic DNA assay. Total TAM production (Table 2.2, ANOVA Fzp = 1.08, P = ), TAMs produced per individual (Table 2.2, ANOVA Fzp = 0.55, P = ) and prevalence of infected strain A individuals (Table 2.2, ANOVA Fzp = 1.01, P = ) did not vary with oligochaete combination. No TAMs were found in any of the control treatments. Table Total triactinomyxons produced per day, infection prevalence and the triactinomyxons produced per infected individual per day. In the 50 strain A + 50 strain B treatment, triactinomyxon production and infection prevalence were calculated based on the fact that only the 50 susceptible Tubifex tubifex (strain A) produced TAMs. Strain B did not produce TAMs. Oligochaete combination Total TAMs per day (±1 SE) Infection prevalence (±1 SE) TAMs produced per infected individual per day (±1 SE) 50 strain A T. tubifex ( ) (±0.10) (±148.59) 100 strain A T. tubifex (± ) (±0.07) (± ) 50 strain A/5 Ostrain B (±1921.5) (±0.04) (±113.18) The range of total TAM production per replicate for each oligochaete combination is as follows: 50 strain A, 382, ,271,757.36; 100 strain A, 366, ,531,636.02; and 50 strain A + 50 strain B, 220, , (Appendix B). A total of 239 progeny were assayed from oligochaete combinations containing strain A and strain B T. tubifex and revealed 2 individuals from the same replicate shared

41 31 bands of both strain A and strain B T. tubifex (Fig. 2.2). Strain A and strain B T. tubifex produced similar numbers of progeny regardless of the oligochaete combination (Appendix C). Figure Polymerase Chain Reaction agarose gel of Tubifex tubifex progeny using randomly amplified polymorphic DNA primer number 2. The progeny were from a replicate exposed to myxospores and containing both strain A and strain B T. tubifex. Al) This row marked with an arrow contains known strain A T. tubifex for identification. A2) This row marked with an arrow contains known strain B T. tubifex for identification. BI) and B2) The rows marked with lightning bolts contain possible hybrid progeny of strain A and strain B T. tubifex.

42 32 Discussion Both strains of T. tubifex exhibited density dependent intra-strain interactions, which is consistent with previous studies (Kosiorek 1974, Bonacina et al. 1989). Strain A T. tubifex exhibited strong intra-strain effects indicated by a decrease in total biomass, adult growth and progeny biomass production as initial density increased. Thus, the intra-strain density dependent effect on strain A total biomass resulted from reductions in both adult growth and reproduction. Strain B T. tubifex also exhibited strong intra-strain effects suggested by a decrease in total biomass and progeny biomass as initial density increased. However, I detected no difference in adult growth of strain B T. tubifex between density treatments, perhaps as result of the high variance of 50 strain B treatments. Thus, the intra-strain density dependent effect on strain B total biomass resulted primarily from a reduction in reproduction. The effects of inter-strain interactions between strain A and B T. tubifex were primarily negative. The total and progeny biomass produced by an individual strain A T tubifex was lowest in the 100 strain A and 50 strain A + 50 strain B treatments. Thus, strain A or strain B T. tubifex have similar effects on the total and progeny biomasses produced by strain A T. tubifex. Adult growth of strain A T. tubifex was similar in the 50 strain A and 50 strain A + 50 strain B treatments and lower in the 100 strain A treatment. Thus, the overall effect is that strain B does not affect strain A adult growth. This

43 33 suggests that the inter-strain density dependent effects on total biomass of strain A occurred because strain B depressed reproduction of strain A. The total and progeny biomass produced by an individual strain B T. tubifex were highest at densities of 50 individuals and lowest in 100 strain B and 50 strain A + 50 strain B treatments. Thus, strain A or strain B T. tubifex have similar effects on progeny biomass produced by strain B T. tubifex. Adult growth of strain B T. tubifex was similar for all treatments and was unaffected by increasing density with addition of strain B or strain A individuals. This suggests that inter-strain density dependent effects on the total biomass production of strain B occurred because strain A depressed the reproduction of strain B T. tubifex. Adult growth of strain A T. tubifex was the only biomass variable that differed among myxospore doses of 0 and Infected strain A adults grew more than those not infected. This result is contrary to a previous experiment where infected T tubifex exposed to the same myxospore dose of 2500 grew less than uninfected T tubifex (Stevens et al. 2001). Two key differences between my study and Stevens et al. (2001) may contribute to the different results. First, Stevens et al. (2001) exposed oligochaetes to myxospores for the duration of the experiment. An extended exposure period to myxospores could have allowed oligochaetes to continue to ingest myxospores thus, having a continuing negative effect on its biomass. Second, the results of Stevens et al. ( ) were reported as the total biomass of adults and progeny of each treatment at the beginning and end of the experiment, as opposed to per capita* day adult growth as I reported.

44 34 An explanation for the higher growth of exposed strain A adults at this particular dose may be parasite manipulation by M. cerebralis. By allocating more energy for growth, strain A individuals may have been manipulated to have larger gut capacity; hence greater TAM production potential. Although progeny biomass production did not differ statistically among myxospore doses, there was a trend of decreasing progeny biomass production as myxospore dose increased suggesting that fewer resources may be allocated to reproductive output due to parasite manipulation. Host manipulations by parasites to increase parasite success have been detected in numerous host-parasite interactions (Moore 1984, Poulin 1994, Thompson and Kavaliers 1994), and include altering host growth to benefit parasite production (Theron et al. 1992). The response of strain B to extremely high doses of myxospores revealed that total biomass and progeny biomasses were higher at a dose of 0 myxospores, and lower at 50,000 myxospores. Adult growth was not affected by myxospore dose. Thus, strain B T. tubifex experiences a fitness cost when exposed to a high abundance of myxospores. This cost could be the result of a debilitating by-product of infection or a damage limitation strategy. Fecundity reduction as a by-product of infection can be a result of host impaired feeding or direct competition between host and parasite for host nutrients (Hurd 2001). Further, damage limitation strategies, such as immune response to tissue damage by infection, can direct resources into tissue repair thereby reducing host fecundity (Hurd 2001). The reduced reproductive output of resistant T. tubifex, consistent with other previous host responses to parasitism, may be a response of

45 35 increased immunity (Minchella 1985, Bayne 1991, Forbes 1993, Sheldon and Verhulst 1996, Stevens et al. 2001). Parasite production and infection prevalence did not differ statistically among oligochaete combinations suggesting resistant strain B individuals did not affect the success of M. cerebralis within strain A individuals. The non-significant trend of deactivation of myxospores by strain B T. tubifex in this research warrants further consideration. A higher dose of myxospores and more replicates per treatment may clarify the relationship between infection prevalence and oligochaete combination. By examining two levels of susceptibility, two relative densities of T tubifex and high myxospore exposure of strain B, I was able to show the effects of density dependent competition and reduced fitness of resistant T. tubifex at high myxospore doses. The intra-strain progeny biomass production of exposed susceptible and resistant T. tubifex were similar and did not decrease with myxospore exposure. However, in previous work by Stevens et al. (2001), susceptible T. tubifex exposed to a dose of 50,000 myxospores had lower total biomass than unexposed individuals. Further, the total biomass and progeny biomass of strain B individuals was low at 50,000 myxospores, suggesting that resistant strain B individuals incur a cost for resistance. The relative costs associated with resistance and infection should be measured. Additionally, although previous model predictions state parasitized, susceptible individuals will bear a disadvantage in competition with resistant individuals (Yan 1996), I was unable to detect this disadvantage. These results may reflect the similarities of physiological characteristics among these two T tubifex strains, such as niche specialization or particle preference that

46 36 create density dependent competition. However, because the strains differed in their ability to propagate the parasite, the results support the idea that the T. tubifex-m. cerebralis relationship differs among T. tubifex strains. Host response to a parasite may be characterized in several ways. A compatible host is one in which the parasite develops and an ensuing cost to the host is incurred, whereas for incompatible hosts, parasites do not develop so there is no cost to the host (Vandame et al. 2000). Compatible hosts have varying degrees of susceptibility with varying levels of parasite development. Resistance by a compatible host prevents parasite development and typically involves an additional cost to the host (Vandame et al. 2000). In general, T. tubifex are compatible with M. cerebralis. Specifically, strain A T. tubifex are susceptible to infection, however it is not clear if strain B T tubifex are less susceptible or resistant to infection, because both responses incur a cost to the host. Host susceptibility varies widely in other host-parasite interactions (Nayar et al. 1992, Ebert 1998, Vandame et al. 2000) and is often coupled with parasite evolution (Ewald 1983, Minchella and Scott 1991). Additionally, parasites may be capable of influencing the evolution of genetic variation for resistance among hosts (Ebert 1998), leading to geographic patterns of adaptation where parasites are most virulent within local host populations and less virulent in novel hosts (Ebert 1994). This suggests that hosts with high susceptibility are local host populations that may have evolved with M. cerebralis, whereas resistant or less susceptible individuals are the novel host. The occurrence of M cerebralis in North America is recent, first documented in 1956 (Hedrick et al. 1998). Because M. cerebralis shows minimal genetic variation among

47 37 isolates (Andree et al. 1999), there is little evidence that co-evolution has occurred. Examining the local adaptation of M. cerebralis and T. tubifex strains will determine the occurrence of such co-evolutionary relationships (Kerans et al. submitted). The discovery of progeny that shared bands of both resistant and susceptible T. tubifex suggests that gene flow among T. tubifex strains may occur at least in the laboratory. However, if the resulting progeny are rare in occurrence due to isolation or reinforcement (Freeman and Herron 2001), then these progeny suggest speciation among T. tubifex strains may be possible. The occurrence of prezygotic or postzygotic isolation of T. tubifex hybrids and their reproductive success must be determined to predict gene flow among strains. If T. tubifex hybrids survive and reproduce successfully, then their ability to propagate M cerebralis must be determined to predict any effects upon hostparasite interactions. I have shown that intraspecific and inter-strain interactions between T. tubifex were density dependent, and the success of the parasite was not reduced by the presence of resistant individuals. Although, resistant individuals do not change infection in susceptible T tubifex, resistant T tubifex still do not produce TAMs. Thus, if the relative frequencies of susceptibility and resistance are known in a specific stream, this could be used to predict salmonid disease severity. For example, if the frequency of resistant individuals was high, we might expect low disease severity in salmonids. However, extrapolating controlled experimental laboratory results to a dynamic natural system may be challenging.

48 38 Future research should clarify the effects of density dependence on strain B individuals (at higher and lower densities), determine the infection potential of hybrid progeny when challenged with myxospores, genetically characterize T. tubifex strains in the two severely whirling disease affected rivers (Madison River MT and Colorado River CO) and finally, determine the historic dispersal of T tubifex strains. These aspects will contribute to whirling disease management decisions and improve our understanding of host-parasite dynamics. My research recognizes the importance of examining host competitive outcomes in host-host-parasite relationships and further characterizes relationships among strains of T. tubifex. Genetic variation of T. tubifex within whirling disease-affected streams is likely influential in the life cycle of this widespread parasite and my results support the recent paradigm in ecology of community genetics that stresses the importance of examining host genetics in host-host-parasite interactions (Whitham et al. 2003).

49 39 CHAPTER 3 OBJECTIVES 2 AND 3: INTERACTIONS AMONG TWO TUBIFICID SPECIES AND A MYXOZOAN PARASITE AND THRESHOLD INFECTION DOSE Introduction Associations between two species, which are often affected by the presence of a third species (Price et al. 1986, Miller and Kerfoot 1987), may influence community structure (Wootton 1994). However, predicting the outcome of these interactions can be difficult because the strength and the direction of the interactions often depend on the environment, life history and population structure of species (Thompson 1988, Power et al. 1996). Although researchers have investigated multiple species predator-prey and competitive interactions (Paine 1966, Schmitt 1987, Kneib 1988, Kohler 1992, Martinez and Medel 2002), studies involving host-parasite interactions are limited (Settle and Wilson 1990, Grosholz 1992, Yan et al. 1998, Bonsall and Hassell 2000). I examined the associations between two aquatic oligochaete species, Tubifex tubifex (Oligochaeta: Tubificidae) and Limnodrilus hoffmeisteri (Oligochaeta: Tubificidae) and a myxozoan parasite, Myxobolus cerebralis (Myxozoa: Myxosporea/ Actinosporea). Myxobolus cerebralis causes whirling disease in salmonids (El-Matbouli et al. 1992) and has caused the declines of at least two populations of wild rainbow trout

50 40 {Oncorhynchus mykiss) within the Intermountain West (USA) (Nehring and Walker 1996, Vincent 1996). The life cycle of the parasite alternates between two hosts: various salmonid species and one Icnown oligochaete species, Tubifex tubifex (Markiw and Wolf 1983, Wolf and Markiw 1984). Myxobolus cerebralis has a complex life cycle with two spore stages, the myxospore and the triactinomyxon (TAM). Tubifex tubifex ingests the myxospore stage while consuming sediments. Parasite reproduction occurs between the epithelial cells of the intestine. After approximately days, depending on T. tubifex strain and temperature (El-Matbouli and Hoffmann 1989, El-Matbouli et al. 1999a, Stevens et al. 2001, Kerans and Stevens in preparation), the buoyant stage of the parasite, the triactinomyxon (TAM), is released into the water. The TAM floats in the water and when a salmonid is contacted, sporoplasms of the TAM are injected into the epidermis. The degradation of the cartilaginous skeletal system occurs as myxospores are produced (El-Matbouli et al. 1995). Myxospores are released into aquatic sediment after infected salmonids die and their carcasses decompose. Benthic assemblages of oligochaetes, which often contain seven or more species (Brinkhurst and Jamieson 1971), may be influenced by intraspecific and interspecific interactions. Tubifex tubifex and Limnodrilus hoffmeisteri, two oligochaetes that typically inhabit stream communities, have similar distributions, physical attributes and functional roles as deposit feeders but different ecological requirements such as particle size and O2 concentration preferences (Brinkhurst and Jamieson 1971, Mermillod-Blondin et al. 2001). In monocultures, individual growth and reproduction of T tubifex and Limnodrilus sp. are density dependent, suggesting intraspecific competition influences

51 41 populations (Kosiorek 1974, Poddubnaya 1980, Bonacina et al. 1989). However, when T. tubifex and L. hoffmeisteri were held in mixed species cultures their respiration rates were lower and population growth rates higher compared to each species in monocultures (Brinkhurst 1974, Milbrink 1987). The authors suggest the interaction between T tubifex and L hoffmeisteri could be a facultative mutualism caused by species providing each other with beneficial food resources (Brinkhurst 1974, Milbrink 1987). It is unknown how M. cerebralis may influence the interactions between T. tubifex and L hoffmeisteri. Myxobolus cerebralis infections are persistent (Gilbert and Granath 2001) and cause reduced growth and reproduction of T tubifex (Stevens et al. 2001). Conversely, L hoffmeisteri cannot be infected by M. cerebralis and its population growth is unaffected by myxospore exposure (Wolf et al. 1986, Kerans et al. submitted). Anecdotal evidence suggests that L hoffmeisteri can consume and deactivate myxospores, which could make myxospores unavailable to T. tubifex (M. El-Matbouli, University of Munich, personal communication), reducing infection potential of T. tubifex. Therefore, an examination of the interactions among susceptible T. tubifex, L hoffmeisteri and M. cerebralis will help determine the effects of mixed species interactions on host and parasite success and contribute to our understanding of the varying effects of whirling disease on salmonids. I conducted two laboratory experiments: the first examined how myxospore availability influenced infection in T. tubifexm d the second examined the interactions among T. tubifex, L. hoffmeisteri and M. cerebralis. The main objectives were to determine: I) the threshold dose of myxospores required to produce 1 0 0% infection

52 42 within T. tubifex, 2) whether the interactions between L. hoffmeisteri and parasitized and unparasitized T. tubifex were similar, and 3) whether L. hoffmeisteri influenced the transmission of M. cerebralis by T tubifex. I expected T. tubifex and L hoffmeisteri would exhibit intraspecific density dependent competition when alone. When together, I expected their interspecific interactions to reflect the facultative mutualism found in previous experiments, hence enhanced growth and reproduction of each species versus that found in monoculture. Further, I expected that the addition of M cerebralis would not affect intraspecific interactions. However, I expected parasitized T. tubifex to have reduced growth and reproduction compared to unparasitized T. tubifex and, with the addition of L hoffmeisteri, T. tubifex would have reduced infection due to deactivation of myxospores by L hoffmeisteri. Methods Several general methods apply to both laboratory experiments that I performed. I used three types of oligochaetes taken from disease-free, pure laboratory cultures, established and maintained since 1997 (see Stevens et al. 2001, Kerans et al. submitted, for details of culture propagation): T tubifex strain A that produced large numbers of TAMs, T. tubifex strain B that produced few TAMs, and L hoffmeisteri that produced no TAMs (Stevens et al. 2001, Kerans et al. submitted). Strain A was collected from the Mt. Whitney California State fish hatchery settling ponds where M. cerebralis was detected

53 43 locally in 1984 (Modin 1998). Strain B was collected from a pond on a tributary to the Gallatin River, Montana, where M. cerebralis had not been detected at collection time (B. L. Kerans, Montana State University, personal communication). Limnodrilus hoffmeisteri originated from the Madison River, Montana and a pond on the campus of Montana State University. Each strain of T. tubifex can be distinguished genetically with molecular markers (Kerans et al. submitted). Oligochaetes were removed from mass cultures and placed in plastic containers for 24 hours without substrate to equalize hunger levels, then moved to plastic containers (9x9x5 cm) with 40 ml of sand, I ml of Organic matter and 200 ml of dechlorinated tap water. I maintained containers, supplied with a constant flow of air, in 15 C incubators on a 12:12 h L:D photoperiod. Oligochaetes were fed 0.10 to 0.12 grams of dried Spirulina weekly, which was sufficient for growth and reproduction (J. C. Lemmon, Montana State University, personal communication). I replaced about 95 % of the water from each container with fresh dechlorinated water weekly. I used oligochaete densities that are common in natural systems (6,000-13,000 individuals/m2) (Lazim and Learner 1986). Experimental densities of 50 and 100 individuals per container translate to natural densities of 6180 and 12,360 individuals/m2. I extracted myxospores from infected rainbow trout {Oncorhynchus mykiss) using a continuous plankton centrifuge (O Grodnick 1975). Trout originated at the Montana State fish hatchery in Ennis MT and were infected in our laboratory using TAMs that originated from T. tubifex cultures of the United States Fish and Wildlife Service, Fish Health Laboratory, Bozeman MT. Using I pi of extracted myxospore emulsion, I

54 44 determined the abundance of myxospores by counting all viable myxospores on a hemacytometer with a compound microscope (400X magnification). The total number of myxospores in the emulsion was extrapolated from the mean number of myxospores from 4 hemacytometer counts and was used to create the required experimental doses. The myxospore containing emulsion or myxospore-free emulsion (derived from parasite-free trout) was added to the containers of oligochaetes. Oligochaetes were exposed to myxospores for 5 days. The 5-day exposure time was based on the average feeding rate of oligochaetes (about ml dry sediment individual'1-day"1 at 15 C) (McCall and Fisher, 1980) and was designed to capture a range of myxospore encounter rates (0.03 through 30.0 myxospores individual"1 day'1) resulting in a range of infection levels. I assumed that all myxospores would be consumed by oligochaetes during this 5-day period based on the feeding rate. After 5 days, all oligochaetes were removed from each container and placed in a new container with fresh substrate and water. Experiment I : threshold infection dose This experiment was designed to determine the threshold number of myxospores needed for 100 percent infection in T. tubifex. I exposed two replicates of 75 oligochaetes of strains A and B to doses of 0, I, 5, 10, 50, 100, 500, 1,000, 5,000 myxospores per oligochaete in a factorial design. On day 89 post-exposure to myxospores, I began to scan for TAMs. The water from each container was poured separately through a 20 pi sieve that retained TAMs.

55 45 Triactinomyxons were back flushed and concentrated in 10 ml of water. I scanned water from two, 55 (J.1 aliquots with a phase contrast microscope (IOOX magnification). I assumed that by day 90 post-exposure all infected oligochaetes would be producing TAMs. On day 90 post-exposure, I took 72 worms from each container and placed them in individual 12-well plates with 4 ml well capacity to determine the prevalence of infected oligochaetes. A day later (after TAMs could accumulate in the wells), I began to scan each well for the presence or absence of TAMs. I scanned each well, using a dissecting microscope (I OX magnification), alternate days, until day 96 post-exposure and recorded wells as positive if TAMs were visible. I calculated the prevalence of infection as the number of oligochaetes producing TAMs divided by the total number of oligochaetes observed. I analyzed the prevalence of infection as a function of myxospore dose using linear logistic regression (SAS Institute, 2001). Because the threshold dose was lower than expected (see Results), I removed the highest myxospore doses from the model to increase its linearity. Experiment 2: interactions between oligochaetes This experiment examined intraspecific and interspecific interactions between T. tubifex (strain A) and L hoffineisteri when oligochaetes were exposed and not exposed to myxospores. I used strain A because it produces many TAMs, thus enhancing my ability to detect the effects of the interactions among the three species. I exposed three replicates of 50 T. 100 T. 50 T A and

56 L hoffmeisteri to three doses of M. cerebralis myxospores (0,10 and 100 per oligochaete) in a response surface competition design (Goldberg and Scheiner 2001). I chose these myxospore doses because they bracket the infection threshold found in Experiment I (see Fig. 3.1), and by doing so, I could observe the effects of interactions among and within species when the prevalence of infection of T. tubifex was at two levels. Additionally, a response surface design allowed me to partition and evaluate intraspecific and interspecific interactions among oligochaetes. For example, comparing response variables between densities of 50 and 100 within a species examines intraspecific interactions and comparing responses between treatments of 100 T. tubifex and 100 L hoffmeisteri and 50 T tubifex + 50 L hoffmeisteri examines interspecific interactions. Starved oligochaetes were wet weighed in groups of 50 to the nearest gram and appropriate numbers of T. tubifex and L hoffmeisteri were placed in prepared containers and inoculated with myxospores. On day 65 post-exposure to myxospores, I removed and weighed all original adults and also separated individuals within 50 T. tubifex + 50 L hoffmeisteri treatments by species according to Kathman and Brinkhurst (1998) using a dissecting microscope. Also at this time, I preserved all progeny in the original sediment with Kahle s solution (Pennak 1978). I counted and weighed progeny at a later date. Once separated, I randomly chose 24 T. tubifex from each replicate to reduce the total number to manageable levels and observed them for infection prevalence, as in Experiment I, and TAM production. Tubifex tubifex were placed in individual 12-well

57 47 plates with water. I extracted all water from each well every other day, replacing it with fresh dechlorinated water. I randomly selected one oligochaete from each replicate to observe individual TAM production for 20 days, thus a total of 3 individuals were observed for each treatments combination. To calculate average TAM production per oligochaete per day, I counted TAMs in two, 55 pi aliquots of water using a phase contrast microscope every three days. Adults from control and L hoffmeisteri treatments were returned to plastic containers after being weighed. I maintained all controls and L hoffmeisteri treatments in incubators and confirmed the absence of TAM production by scanning water from the experimental containers as described previously in Experiment I. The effects of host-parasite interactions are often associated with reduced host growth and reproductive success (Forbes 1993). I measured three responses to assess the effects of the interactions among oligochaetes and M. cerehralis on oligochaete fitness: I) total biomass production per initial individual (((final biomass of adults-initial biomass of adults) + progeny weight) / initial density of adults) (hereafter referred to as total biomass ), 2) growth rate of adult oligochaetes per initial individual ((final biomass of adults-initial biomass of adults)/initial density of adults) (hereafter referred to as adult growth ) and 3) progeny biomass produced by an individual adult (final progeny biomass / initial density of adults) (hereafter referred to as progeny biomass ). I separated total biomass into two contributing response variables of adult growth and progeny biomass to determine how each factor influenced total biomass. I did not examine oligochaete mortality as a response variable because significant mortality typically occurs only when

58 48 myxospore doses are extremely high (Stevens et al. 2001). Oligochaete mortality during the experiment was 7 % (±1% SD). Oligochaete species-density combination will be further referred to as oligochaete combination. I tested total biomass, adult growth, and progeny biomass of each oligochaete species for significant differences among oligochaete combinations and myxospore dose using two-factor analyses of variance (ANOVA, SAS Institute, 2001). I examined the residuals for normality and stabilized the variances with log or square root transformations, where necessary. I also tested the prevalence of infected T. tubifex and average TAM production for significant differences among oligochaete combinations and myxospore dose using two-factor ANOVA, which required square root and log transformations, respectively, to stabilize the variances. Significant main effects of ANOVA s were followed by Tukey-Kramer Honestly Significant Difference test (Tukey s HSD) to determine which levels of main effects differed from each other. Results Experiment I : threshold infection dose I observed 1152 T tubifex for the prevalence of infection for two weeks. The prevalence of infection of strain A reached 100% at a myxospore dose of 50 per oligochaete, with a dramatic rise between dosses of 5 arid 50 myxospores per oligochaete (Fig. 3.1, Appendix D). The linear logistic regression was not significant when all points

59 49 were included (Fi, 13 = 1.96, P = ). However, regression was significant (Fk9 = 14.27, P = 0.004) when infection prevalence results of doses at or below 100 myxospores were used. I detected no TAMs produced by strain B (Appendix D) I. S S I 2d IE O O Strain A \ 7 Strain B 0 I Dose of myxospores Figure The relationship between myxospore abundance and the percentage of Tubifex tubifex releasing triactinomyxons. Experiment 2: interactions between oligochaetes The total biomass of T. tubifex was highest at densities of 50 individuals and lowest at densities of 100 individuals of 7. tubifex or 50 T. tubifex in combination with 50 L. hoffmeisteri (Table 3.1, Tukey s HSD, P < 0.05, Fig. 3.2a). Total biomass did not differ between myxospore doses, or with the interaction between oligochaete combination and myxospore dose (Table 3.1, Fig. 3.2a).

60 50 Growth of adult T. tubifex was highest at densities of 50 individuals alone or in combination with 50 L hoffmeisteri and lowest in 100 T. tubifex treatments (Table 3.1, Tukey s HSD, P < 0.05, Fig. 3.2b). Growth of adults was greater in the 10 myxospore treatment than in the no myxospore treatment (Table 3.1, Tukey5s HSD, P < 0.05). All other myxospore dose comparisons were not significant. The interaction between oligochaete combination and myxospore dose was not significant (Table 3.1). Tubifex tubifex produced greater biomass of progeny at densities of 50 individuals alone, but as densities increased to 100 with the addition of T. tubifex or L. hoffmeisteri progeny biomass decreased (Table 3.1,Tukey5S HSD, P < 0.05, Fig. 3.2c). Progeny biomass did not differ among myxospore doses, or with the interaction between oligochaete combination and myxospore dose (Table 3.1). The total biomass of L. hoffmeisteri was highest at densities of 50 individuals and decreased as densities increased to 100 with the addition of L. hoffmeisteri or T. tubifex (Table 3.1, Tukey5s HSD, P < 0.05, Fig. 3.2d). Total biomass did not differ among myxospore doses, nor with the interaction between oligochaete combination and myxospore dose (Table 3.1, Fig. 3.2d). Growth of adult L. hoffmeisteri did not vary among oligochaete combinations' (Table 3.1, Fig. 3.2e). However, growth of adults was highest in the 100 myxospore treatments and lowest in the no myxospore treatments (Table 3.1,Tukey5S HSD, P < 0.05, Fig. 3.2e). The interaction between oligochaete combination and myxospore dose was not significant (Table 3.1).

61 51 Table Analysis of variance results on the effects of oligochaete combination and myxospore dose for biomass response variables. All biomass response variables are expressed as a per individual raie. Response Oligochaete variable snecies* (T) Source of variation SS F df P Total biomass Tt dose oligochaete combination interaction error ' Lh (sqrt) dose oligochaete combination interaction error Adult growth Tt dose oligochaete combination interaction error Lh dose oligochaete combination interaction error Progeny biomass Tt (sqrt) dose oligochaete combination interaction error Lh (sqrt) dose oligochaete combination interaction error (f) = The transformation (sqrt)=square root and (In)-Iog. ss = sums of squares.

62 52 Tubifex tubifex Limnodrilus hoffmeisteri S = c. O alone, O myxospore D w/ LH, O myxospore CD alone, IO myxospore 8 cd w E E3 w/lh, 10 myxospore!"i 15 alone, 100 myxospore w /LH, 100 myxospore -o 13 C) O alone, O myxospore D w/ TT1O myxospore O alone, 10 myxospore EU w/ TT, 10 myxospore # alone, 100 myxospore w/tt, 100 myxospore S r. 5 O I I Oligochaete combination Oligochaete combination Figure The effects of oligochaete species and density and M cerebralis infection on biomass variables. Each column depicts the particular variable for the target species listed, a) per capita biomass, b) per capita adult growth, c) progeny biomass produced per capita, d) per capita biomass, e) per capita adult growth, f) progeny biomass produced per capita. Error bars are one standard error of the mean. Ifoverlap of individual biomass values occurred on the graph, points were offset.

63 53 Limnodrilus hoffmeisteri produced greater biomass of progeny at densities of 50 individuals, and as densities increased to 100 with the addition of T. tubifex of L hoffmeisteri progeny biomass decreased (Table 3.1, Tukey s HSD, P < 0.05, Fig. 3.2f). Progeny biomass did not vary among myxospore doses or with the interaction between oligochaete combination and myxospore dose (Table 3.1). I observed 648 individual T. tubifex for infection prevalence (24 individual T. tubifex from each replicate containing T. tubifex). The infection prevalence was lowest for 100 T tubifex treatment, and highest in 50 T. tubifex, and 50 T tubifex + 50 L hoffmeisteri treatments (ANOVA: F 2,12 = 7.50, P = 0.007, Tukey s HSD, P < 0.05, Fig. 3.3). (Z) 10 myxospores/worm myxospores/worm Figure The effect of oligochaete density and composition on the proportion of Tubifex tubifex releasing TAMs (N=3). Error bars represent one standard error of the mean. Bars that share the same letter do not differ significantly (P < 0.05) based on Tukey s HSD test.

64 54 Further, the prevalence of infection increased as myxospore dose increased (ANOVA: F 1,12= 25.85, P = , Tukey s HSD, P < 0.05, Fig. 3.3). The interaction between oligochaete combination and myxospore dose was not significant (ANOVA: F 2,12 = 0.65, P = 0.53). I observed 18 individual T. tubifex for TAM production (one individual from each replicate exposed to myxospores)(fig. 3.4). Production of TAMs by an individual did not differ with oligochaete combination (ANOVA: F 1,12= 0.71, P = 0.414), myxospore dose (ANOVA: F 2,12 = 0-69, P = 0.518), or the interaction between the two (ANOVA: F 2,2= 1.74, P = 0.217). Treatments of 50 and 100 L hoffmeisteri and negative controls did not produce TAMs. The triactinomyxons produced by an individual T. tubifex for each oligochaete combination range from , for 50 T. tubifex', , for 100 T tubifex', and 1, , for 50 T tubifex with 50 L hoffmeisteri (Appendix E). If Ij I a % IZI I myxospores/worm 100 myxospores/worm 50 T tubifex 100 T. tubifex 50 T. tubifex + 50 L hoffmeisteri Oligochaete combination Figure The average TAM production of individual Tubifex tubifex over a 20-day period (N=3). The error bars represent one standard error of the mean.